considerable advantages, too, for the spread of producer co-operatives. In addition, while bypassing the obstacle posed by economic fragmentation, such investments would nevertheless be attacking it, thus raising the degree of economic integration. No doubt, these investments would require industrial inputs at a higher level than before and the financing of this might imply that the planners have to give up some of the surplus extracted from the agricultural sector for use by it within its boundaries. In our opinion, this approach provides the basis for achieving high growth targets in the medium term without compromising on the distributional front at the class, sector, or regional levels. Two qualifications need to be registered. First, this does not imply that the DTYP target of y = 7.5 per cent per annum becomes feasible in this strategy. Even in Case A, the argument was only partly that it was probably not achievable; rather, that achieving it with n = 3.5 per cent would almost certainly lead to a vicious inflationary spiral, thereby worsening income distribution. In Case B, the burden of financing would be shared in an egalitarian manner through the rationing system but its average level would not be any different. What is being argued is that, first, for any given n, y* (C) > y* (A, B), and second, the rate of growth of n would be substantially greater over time in Case C than in Case A or B. Thus, Case C could be viewed as laying the basis for an eventual second phase of an industrialisation drive of the type now being proposed, in our view, prematurely. Second, it is probable that under Case C, rural foodgrain consumption would rise in the short run. In this strategy, too, state farms would play a crucial part in the transitional phase and beyond. It is necessary therefore to assist them in achieving efficiency quickly, and to overcome the problems of haphazard location and early growth. A period of consolidation might be necessary prior to any further expansion on any large scale. Finally, we need to turn our focus to the problems of urban poverty and unemployment which are not directly handled in any of the three cases. A separate policy component is therefore called for. A two-pronged approach is necessary. The first of these is to ensure that all low-income earners are covered by the urban rationing system. In the present context, this would require extending the coverage to the smaller urban centres and even in the larger ones to that lowest strata which might not be registered in any urban kebele. Thus, the AMC needs to grow greatly and quickly. It is in this context that the current and future role of the state farm sector has to be seen. Even within the framework of Case C, it will be some time before the area of stable grain yields is extended to a point where the urban populations are not held to ransom by the weather all too frequently; in the meantime, the state farms provide an insurance cover which is indispensible. (A corresponding function would be performed in the food-insecure rural areas by the grain banks suggested earlier.) Further, the kebele shops need to move more into the inferior cereals, in particular, sorghum, maize and black teff. Improving the storage facilities of the AMC and state farms could achieve the welcome result of lowering cost by anything up to 15-20 per cent on some crops. All such gains registered should reflect themselves in lower prices for the inferior, rather than for the superior, cereals as appears to have been the case in the recent past.